This is a project to develop new agents for administration in the pre invasive stage of lung cancer. Vitamin A analogues have the greatest reported activity of any of the agent used in chemoprevention of tobacco-related malignancies. The relationship between the administered dose and the resulting in vivo dose at the site of the target organ is poorly understood. We propose to change the route of administration of the retinoid to improve the therapeutic index. This work has been supported by preclinical toxicology studies conducted under a NIH CRADA with Battelle Labs. The amount of drug required to saturate the bronchial epithelium when delivered as an aerosol is significantly lower than the systemic dose required to deliver a comparable amount of drug to the target organ. Early lung cancer exists only in the tracheobronchial tree where the effects of chronic cigarette smoke have caused diffuse cancerous injury. The beneficial effect of retinoids on cancer cell growth is a direct drug/cancer cell interaction. If the airway is effective as a conduit for the carcinogen, it may also be the most effective way of directing retinoids to the affected early lung cancer cells. Aerosolizing retinoids into the airway instead of giving them orally can result in better bioactivity of the chemoprevention agent while potentially causing less frequent systemic side effects since the blood levels after an aerosolized dose will be much lower. We are proceeding to conduct an initial Phase I/II trial of aerosolized retinoids at the Clinical Center. Two papers outlining the success of this effort in tobacco-carcinogen induced lung cancer models in rodents have been published. Due to the substantial progress on this effort, we will be able to start a clinical trial in late 2000. A range of other active cytotoxic agents will also be evaluated to determine if limited intrapulmonary drug exposure is sufficient to control early, small volume lung cancer. A Phase I trial with aerosolized doxirubicin is ongoing which represents a first step in this regard. With recent developments with spiral CT, routine detection of such small primary lesions is expected and this may drive the need for effective aerosolized chemotherapeutic drugs. The preclinical rationale for this approach has been bolstered by a large in vivo experiment with carcinogen-induced lung tumors. In a series of exposures through a range of retinoid doses, a clear dose response emerges as to the success of retinoid to suppress the frequency of lung tumors developing. We recently completed a Phase II clinical trial which evaluated the efficacy of a locally administered pan cyclooxygenase inhibitor in a randomized trial of individuals with oral leukoplakia. The report has been accepted for publication in the journal Clinical Cancer Research. The cyclooxygenase inhibitor was well tolerated but did not lead to enhances drug response when compared to placebo control. As discussed in the report, this experience has given us important feedback to consider in the development of subsequent drug delivery efforts for early epithelial cancer. A challenge with this approach is that the pharmaceutical industry does not anticipate that a large enough target population will exist to justify the requisite investment to develop these drugs. We have worked with the radiology community to integrate the development of these tailored delivery approaches with the types of early cancer patients being defined by spiral CT screening to determine if this raises the priority of pharmaceutical participation so this new class of public health tools can be developed.